ELT News, June 2010

Oral Mediation In The KPG Exams

Although the concept of ‘mediation’ is not new,[1] the inclusion of mediation activities in an exam battery is an innovation that the KPG exams introduced in the field of Language Testing. Last month, the ‘KPG Corner’ hosted an article which explains how the concept of mediation is viewed within the context of the KPG exams, but the focus is on written mediation and mediation strategies that might lead to successful performance. This article will be concerned with oral mediation and the characteristics of the mediation activities included in the oral test of the KPG exams from level B1 onwards.  

Oral mediation task demands across levels

Independent of the level of proficiency for which they are designed, all written and oral mediation activities have one common feature: a source text in Greek from which candidates extract information and relay it in English. However, whereas in written mediation activities the source and the target text are of the same channel of communication, in oral mediation the channel of communication is different. The task input is in the form of a written (Greek) text, whereas the output is in the form of oral speech in English.

The output expected at each level of production differs, but there is also a substantial difference between task demands of activities in B and C level oral tests. This is because B level oral mediation tasks require one sided talk, whereas C level tasks require exchange of information between two speakers –two candidates. The different expectations and oral task demands across levels of proficiency are recorded systematically thanks to a Task Analysis Project that the RCeL research team has been carrying out since 2007. The aim of this project is to analyze the tasks designed for all the modules and levels in all the KPG languages. For the purposes of the project, a Task Description Model (TDM) has been created. Its basic categories of analysis are: Topic, Genre, Generic Process, Speaker-Audience Roles and Lexicogrammar.[2] The TDM has been used for the analysis of the oral mediation tasks by Nteliou,[3] and some of the findings, relevant to this brief article, will be presented.

B1 level

Concerning the B1 level oral mediation, the prompts used are Greek texts of about 150-200 words each and the goal of the activity is to assess candidates’ ability to talk in English about the topic of the Greek text, by relaying in the target language the main or specific points contained in the source text. Sometimes, there might be two or more short Greek texts and candidates may be asked to compare the information therein.TDM analysis of the B1 level activities reveals that source texts consistently offer advice or provide instructions, in bullet form or in separate short sentences. Fewer source texts provide a description or contain factual information. The text topics usually draw from everyday, familiar situations (e.g. family life, school, entertainment) and their lexicogrammar is quite simple (Nteliou, 2010). Target texts –i.e. the speech that candidates are required to produce– also offer advice, provide factual information or give instructions. Actually, analysis has shown that the communicative purpose of most target texts at this level is to give instructions or advice (Nteliou, ibid). This means that the communicative purpose of the source and the target texts is often the same. However, because the channel of communication in each instance is different, as are the contextual features of the source and target text, the output generated is expected to be distinct. That is, the speech that candidates are to produce is not to be a parallel to the text they draw the information from. Put differently, the output expected is not a translation of the source text. The task asks candidates to use some of the ideas from the source text and produce one of their own, for reasons the task rubrics specify. For example, the source text may be 4 postcards sent by a Greek family on vacation. The task asks the candidate to skim through them and to inform the examiner where the Greek family is, what they’re doing or how they’re spending their time, how long they’ll be staying, etc. 

Demands regarding the quality of output are predetermined by the test team (see Karavas: 2009: 7). They are no different from those in other oral production activities, discussed in the January issue of the ELT News.[4]

B2 Level

The B2 level oral mediation tasks are designed on the basis of the same rationale and goal as those at B1 level, although they are linguistically and cognitively more demanding. In other words, candidates are asked to extract information from the Greek source texts (200-250 words in all) and relay it in English, for a reason specified in the task rubrics. The communicative purpose of source texts varies. Many of them offer advice or instructions as at B1 level, but often advice is also accompanied by descriptions or explanations (Nteliou, ibid). The topics of the source text may refer to everyday life and familiar matters, but they may also be on special issues, such as the benefits of music or why water is good for us. Here is an example:

Task rubrics are usually longer than at B1 level, and may ask that candidates give advice to a friend, explain something to someone or do several things at the same time: e.g. inform, describe and invite. For example, a task motivated by the source text above may be:

Imagine I am your friend’s 12 year old daughter (or son) daughter, and that I want to know why this text has a picture of someone drinking water, and what this text is about. Inform me and then tell me why it’s important for me to drink lots of water during the day.As has already been said above, demands regarding the quality of output are predetermined by the test team (see Karavas: 2009: 7).

When reading through the assessment criteria for this level,[5] one will note that besides demands as to the accuracy of the lexicogrammar of the speech produced, candidates are also are expected to be more fluent, with fewer hesitations and pauses than at B1 level, and the speech they produce should be appropriate for the context of situation. This means that candidates are expected to carefully consider who they are talking to and why, so as to use the proper register and use the appropriate language. For example, in the example above, their speech should be adjusted to their imaginary interlocutor, i.e. a 12 year child.

C1 Level

The oral mediation activity at this level is different from B level activities. First of all, the couple of candidates being tested are provided with different texts on the same issue and assigned the same task, which required them to exchange information. The examiner assigns the task and does not participate in the conversation at all.

The goal of this activity is similar to that of the other levels insofar as candidates have to read a Greek text (or texts) and then relay specific information in English. However, at this level, candidates are required to converse with one another since they have to exchange information they find in the Greek source text and then reach a common decision, or find a common solution to a problem. In other words, C1 level tasks involve candidates in a negotiation process. They negotiate information and also which is the best choice, solution to a problem, etc. For example, a task based on two different leaflets each one of which contains book announcements, could be the following:

Exchange information with your partner about the books in your leaflet and together decide which one is likely to interest young people and which two are most likely to interest male and female adults.

The candidates’ ability to converse is evaluated. Actually, in addition to the criteria used for the B level activities,[6] C1 level candidates are also graded both for their interaction and their mediation skills.[7] These two additional criteria clearly show concern that the C level candidate be able to initiate, sustain and end a conversation, as well as to selectively relay information form the Greek text, without adding irrelevant information or resorting to translation.

In terms of language performance, C1 level candidates are often expected to use some technical vocabulary that relates to the topic at hand, and to use sophisticated communication strategies to overcome these gaps of knowledge. In general, they are expected to be rather fluent and to use language which is correct and appropriate.

How to prepare candidates for the oral mediation activities

Although oral mediation is something that any foreign language speaker is involved in when in daily communication with others, preparation though practice is likely to lead to more successful results. It is important that candidates become familiar with the test format and task demands as with any test activity. In the case of oral mediation, it is even more important to train candidates in the use of specific strategies which will help them read the source text with the specific purpose of extracting precise information, and to relay it in a way that is relevant to the context of communication, as set by the task. This is to say, successful performance in mediation actually requires the development of unique socio-cognitive skills, which some candidates may have developed outside of the foreign language classroom but others may not have. For that reason, it is important for those teachers who are preparing candidates for the KPG exams to provide opportunities for training to take on the role of mediator –a role we have to take on in our daily lives, as speakers of two or more languages.

References

Council of Europe . 2001. Common European Framework of Reference for Languages. Learning, teaching,  assessment. Cambridge: Cambridge University Press.

Dendrinos, B. 2006. “Mediation in Communication, Language Teaching and Testing.” Journal of Applied Linguistics, 22, pp. 9-35.

Karavas, K. 2009. The KPG Speaking Test in English: A Handbook. Athens: RCeL Publications, University of Athens .

Κονδύλη, Μ. & Λύκου, Χ. 2009. ‘Γλωσσολογική περιγραφή των θεμάτων του ΚΠΓ: Η οπτική του κειμενικού είδους και της λεξικογραμματικής’. RCeL Research Periodical, p.1- 17: http://rcel.enl.uoa.gr/periodical/article_periodical_01.pdf

Nteliou, Ε. 2010. “Oral task characteristics and task performance: in search of patterns”. Paper presented at the second Postgraduate Conference of the Faculty of English Studies, University of Athens, 16/01/2010.

Elfie Nteliou and Bessie Dendrinos